1. We are doing so to have control over the methods. We usually want a method X to be executed for N seconds and be terminated after that. The same thing can be better achieved with a combination of threads and multiprocessing.
2. We can make use of the multiprocess.pool library to see whether we get better results and, furthermore, increasing the processor cores is always a good option.
3. Yes, there is another Python-Nmap utility, called libnmap: https://libnmap.readthedocs.io/en/latest/process.html.
4. Yes, we can. Please explore more on this.